The Match IQ capability primarily focused on assisting data stewards with post-match entity resolution.
A few key points:
- Match IQ models run in parallel to existing rule-based matching and appear as a separate "Match Model" that stewards can enable.
- Users can configure actions like auto-merge, review, or leave separate based on model confidence scores.
- The demo shows workflows of stewards reviewing and validating potential matches surfaced by the ML models.
- There was discussion around how models could learn from stewards' merge/unmerge actions to improve over time.
- The overall goal appeared to be reducing manual effort for stewards when resolving duplicate entities.
So in summary, Match IQ seems optimized for consumption by data stewards doing entity resolution, rather than for developers building the underlying match rules. The models aim to augment the existing rule-based approaches in the system.
------------------------------
Suchen Chodankar
------------------------------
Original Message:
Sent: 07-28-2023 15:23
From: Chris Detzel
Subject: Is the new Match IQ capability for data stewards for post match entity resolution or is this for developers to create the match rules?
More information on this would be helpful
------------------------------
Chris Detzel
Director of Customer Community and Engagement
Reltio
------------------------------